Science cloud

Scientific service cloud

Our computing infrastructure based on cloud computing separates functionality provided by services from the underlying hardware. Services we provide can run on our own hardware, or are delivered as images that can be run on any cloud infrastructure, internal or public, commercial or free.

By implementing services managed by an Enterprise Service Bus as virtual applications running on a cloud, the computing needs of an experiment can grow or shrink dynamically as needed, and fully utilise available hardware.

Services can be used standalone, in parallel, and through the Minerva modelling and inference framework.


Advantages

  • Wrap legacy codes in modern services deployed in the cloud
  • Flexible remote calling from virtually any programming/script language
  • Combine services to form large integrated models
  • Virtualization technology creates independence from hardware, and optimal scaling of computing resources
  • Use together with the Minerva framework for powerful model construction and inference using legacy code.
  • Enterprise service bus provides generic front end and flexible management, load balancing etc, of an experiment’s services.